Adaptive Secrecy in the Making of the Atomic Bomb: Toward a Process View of Secretive Innovation

Secrecy is critical for knowledge protection during innovation, but it can hinder knowledge creation. Despite considerable research into when and why organizations employ secrecy, our understanding of how secrecy is managed during the research and development (R&D) process remains limited. I examine how secrecy emerges, manifests, and evolves across the R&D process through a historical case study of the making of the atomic bomb in the United States during World War II. I uncover a model of adaptive secrecy, whereby practices of knowledge sharing and concealment evolve with changing tensions between knowledge creation and protection. Adaptive secrecy practices give rise to different types of uncertainty (evaluative, boundary, and performance) that are navigated through three types of adaptive disclosures: revealed concealment, revealing meta-information, and revealing contextual information. Through this process, organizations develop complex, multilayered secrecy structures with in-groups, out-groups, and in-between groups. I articulate feedback relationships between elements of the adaptive secrecy model and establish boundary conditions pointing to research directions on variants of this process. This study contributes a dynamic, microfoundational view of secretive innovation, develops the notion of out-groups and in-between groups within organizations, and advances the idea of meta-knowledge sharing in collaborative work. Supplemental Material: The online appendix is available at https://doi.org/10.1287/orsc.2023.17687 .